ABSTRACT

Title of Document: ENHANCED MATURITY

MODELING ANALYSIS FOR

SUSTAINABLE CONCRETE

MIXTURES USING HIGH VOLUME

FLY ASH.

Shardul Pendharkar, M.S Civil &

Environment Engineering, 2015

Directed By: Associate Professor, Dr. Dimitrios

Goulias, Civil & Environmental

Engineering Department

Construction industry is reluctant to use high volumes of fly ash in

concrete due to slower rate of strength gain observed in high volume fly

ash concretes. Fly ash, a by-product of coal industry, is readily available

on large scale, has cementitious properties and can be used as cheaper

substitute for cement. In case of large concrete structures, a phenomenon

known as mass effect is observed which cures the concrete at higher

temperature thereby accelerating the rate of strength development in early

ages. The aim of this research is to demonstrate that, although the rate of

strength gain is slower in HVFA concrete, sufficient early age strength is

developed due to mass effect observed in large concrete structures. A 14-

day maturity based approach was adopted to develop prediction models

which can estimate in-place strength of HVFA concrete mixtures by

taking into account the mass effect of concrete. The results maturity are

compared against the existing methods and also with 28-day based

maturity models and attempt is made to select the best approach of in-

place strength determination.

ENHANCED MATURITY MODELING FOR SUSTAINABLE CONCRETE

MIXTURES USING HIGH VOLUME FLY ASH.

By

Shardul Pendharkar.

Thesis submitted to the Faculty of the Graduate School of the

University of Maryland, College Park, in partial fulfillment

of the requirements for the degree of

Master of Science

2015

Advisory Committee:

Dr. Dimitrios Goulias, Chair

Dr. M. Sherif Aggour

Dr. Amde M. Made

© Copyright by

Shardul Pendharkar

2015

ii

Preface

A major decision in construction industry today revolves around

predicting the in-place strength of concrete and the timely removal of

formwork. The removal of formwork can be facilitated when the concrete

develops sufficient strength for the purpose for which it is cast. Due to its

slower rate of strength gain, the construction industry is reluctant to use

high volume fly ash (HVFA) concrete. Past research indicates that the

slower rate of strength gain can be offset by the high in-place hydration

that takes place inside large concrete structures. The aim of this research

is to develop 14-day based maturity models to capture this high in place

hydration and develop prediction models to estimate in-place strength of

concrete mixtures. Four types of mixtures were analyzed in this research:

one conventional concrete mixture, concrete mixture with 35% of type F

fly ash, concrete mixture with 50% of type F fly ash and concrete mixture

with 35% of type C fly ash. Three different approaches to calculate

activation energy were considered and prediction models based on each

approach were developed. Each approach was based on 14-day data. The

first approach was based on ASTM C1074, also known as variable Su

approach. In this approach, the limiting strength of each mixture was

considered different at different curing temperatures. The second was

iii

based on setting time of each mixture and is known as setting time

approach. The third approach was known as constant Suc approach. In this

approach, the limiting strength for each mixture was kept constant

irrespective of curing temperature. The results of in place strength

estimated from the 14-day maturity based prediction models of each

mixture were compared with the results from the existing methods of in

place strength determination like field cured method, pullout testing,

match curing and the 28-day based maturity models available from

previous phase of research. In pullout testing, a metal insert is embedded

in concrete at specific locations and the force required to pull it out is

determined. This force is correlated with the compressive strength of

concrete. In match curing, the actual temperature profile of concrete

structure is replicated in testing cylinders through the use of temperature

sensors and micro-controllers. The pullout force and compressive

strength correlations developed in this research were based on 14-day

data. Based on this research, the following primary conclusions were

reached: the 14-day maturity models show improvement in estimating the

in-place compressive strength for fly ash of type F over the 28 –day

maturity model; the 14-day setting time approach that was explored in

this research has shown encouraging results and this approach can be

iv

used for developing prediction models; the 14-day setting time method

and 14-day maturity method consistently provide lower errors of

prediction errors than the 28-day maturity models for all the HVFA

concrete mixtures.

v

Dedication

To my late grandparents, Mr. Vasudeo Pendharkar and Mrs. Sudha

Pendharkar

vi

Acknowledgements

I would sincerely like to thank my research advisor, Dr. Dimitrios

Goulias, under whose constant guidance and support, this research was

possible. The opportunity of working with Dr. Goulias over the course of

this research has truly enriched me technically, personally and

professionally. For that, I am deeply indebted to him.

I would also like to express my gratitude to Dr. Aggour and Dr. Amde for

agreeing to be a part of my defense committee. Their quick and timely

responses to my concerns helped smoothen the process of thesis defense.

I thank my parents and brother for their unwavering support and

unconditional love. Their continued faith has always inspired and

motivated me. Lastly, I want to thank all my friends and other family

members who have stood by me during thick and thin.

vii

Table of Contents

Chapter 1 Introduction .................................................................................................. 1

1.1 Research Outline ........................................................................................... 1 1.2 Research Methodology ....................................................................................... 2

1.2.1 Activation Energy ........................................................................................ 4 1.2.2 Maturity Modeling ....................................................................................... 5

1.2.3 Match Curing ............................................................................................... 5 1.2.4 Field Curing ................................................................................................. 7 1.2.5 Pullout Tests................................................................................................. 7

1.3 Organization of Thesis ........................................................................................ 8

Chapter 2 Background ................................................................................................ 10 2.1 Development of Maturity Functions ................................................................. 10 2.2 Strength-Equivalent Age Relationship ............................................................. 15

Chapter 3 Materials & Concrete Mix Design ............................................................. 18 3.1 Raw Materials ................................................................................................... 18

3.2 Mixture Proportion............................................................................................ 21 Chapter 4 Experimental Work, Analysis and Results ................................................. 24

4.1 Determination of Activation Energy ................................................................. 24

4.1.1 Activation Energy by ASTM C1074 ......................................................... 24 4.1.2 Activation Energy by Setting Time Method .............................................. 39

4.1.3 Activation Energy by Constant Suc Approach ........................................... 48 4.2 Actual in-place strength due to mass concrete effects ...................................... 50

4.2.1 Block Testing ............................................................................................. 52

4.2.2 Slab Testing ............................................................................................... 55

4.3 Pullout Testing .................................................................................................. 56 4.3.1 Compressive Strength vs Pullout Force Correlation .................................. 57 4.3.2 Pullout Tests on concrete blocks................................................................ 64

4.3.3 Pullout testing on concrete slabs ................................................................ 64 4.4 Field Curing ...................................................................................................... 67

Chapter 5 Exploring Alternative Approaches for Maturity Modeling........................ 70

Chapter 6 Summary, Conclusions and Recommendations ......................................... 97

viii

List of Tables

Table 3.1: Gradation of Coarse and Fine Aggregates ................................................. 19 Table 3.2: Properties of Coarse and Fine Aggregates ................................................. 20 Table 3.3: Mix Design Specifications for Mortar Mixes ............................................ 22 Table 3.4: Mix Design Proportions for Concrete Cylinders ....................................... 23

Table 4.1: Setting Time for Mortar Mixtures (ASTM C403) ..................................... 27 Table 4.2: 28-day vs 14-day k value comparison ....................................................... 28 Table 4.3: Activation Energy (ASTM C1074) Based on 14 days data ....................... 32 Table 4.4: Compressive Strength and Equivalent Age for Control Mix Based on 14

days data...................................................................................................................... 33

Table 4.5: Compressive Strength and Equivalent Age for 35% FA-A Mix Based on

14 days data................................................................................................................. 33

Table 4.6: Compressive Strength and Equivalent Age for 50% FA-A Mix Based on

14 days data................................................................................................................. 34 Table 4.7: Compressive Strength and Equivalent Age for 35% FA-C Mix Based on 14

days data...................................................................................................................... 34

Table 4.8: Activation Energy by Setting Time Approach .......................................... 41 Table 4.9: Compressive Strength and Equivalent Age for Control Mix based on

Setting Time Method .................................................................................................. 41

Table 4.10: Compressive Strength and Equivalent age for 35% FA-A mix based on

Setting Time Approach ............................................................................................... 42

Table 4.11: Compressive Strength and Equivalent Age for 50% FA-A Mix based on

Setting Time Method .................................................................................................. 42 Table 4.12: Compressive Strength and Equivalent Age for 35% FA-C Mix based on

Setting Time Method .................................................................................................. 43

Table 4.13: Rate Constants using Constant Suc Approach .......................................... 49 Table 4.14: Activation Energy by Constant Suc Approach ......................................... 50 Table 4.15: Match Cured Strength-Control Mix Block .............................................. 54

Table 4.16: Match Cured Strength-35% FA-A Mix Block......................................... 54 Table 4.17: Match Cured Strength-50% FA-A Mix Block......................................... 54

Table 4.18: Match Cured Strength-35% FA-C Mix Block ......................................... 55 Table 4.19: Match Cured Strength-Control Mix Slab ................................................ 56 Table 4.20: Match Cured Strength-50% FA-A Slab ................................................... 56 Table 4.21: Pullout Force and Compressive Strength for Control Mix ...................... 58

Table 4.22: Pullout Force and Compressive Strength for 35% FA-A Mix ................ 58 Table 4.23: Pullout Force and Compressive Strength for 50% FA-A Mix ................ 59

Table 4.24: Pullout Force and Compressive Strength for 35% FA-C Mix ................. 59 Table 4.25: Pullout Force and estimated Compressive Strength for Control Mix Block

..................................................................................................................................... 65 Table 4.26: Pullout Force and Compressive Strength for Control Mix Slab .............. 65 Table 4.27: Pullout Force and Compressive Strength for 35% FA-A Block.............. 65

Table 4.28: Pullout Force and Compressive Strength for 50% FA-A Block.............. 66 Table 4.29: Pullout Force and Compressive Strength for 50% FA-A Slab ................ 66 Table 4.30: Pullout Force and Compressive Strength for 35% FA-C Block .............. 66

Table 4.31: Field Cured Strength for Control Mix Block ........................................... 67

ix

Table 4.32: Field Cured Strength for Control Mix Slab ............................................. 68

Table 4.33: Field Cured Strength for 35% FA-A Mix Block ..................................... 68 Table 4.34: Field Cured Strength for 50% FA-A Mix Block ..................................... 68 Table 4.35: Field Cured Strength for 50% FA-A Mix Slab ........................................ 69

Table 4.36: Field Cured Strength for 35% FA-C Mix Block ..................................... 69 Table 5.1: Comparison of Match Cure Strength with Predicted Strength from the 14-

day and the 28-day Pullout Correlations ..................................................................... 73 Table 5.2: Percent strength prediction differences between match cured versus the 14-

day and the 28-day Pullout correlations...................................................................... 74

Table 5.3: Strength Prediction Comparison between Maturity Model based on 14-Day

and Pullout Test .......................................................................................................... 78 Table 5.4: Percent Difference between in predicted strength between 14-day pullout

and 14-day maturity method ....................................................................................... 79

Table 5.5: Strength Comparison by Different Methods.............................................. 85 Table 5.6: Percent differences between match cured and other methods ................... 86

Table 5.7: Error Analysis for the Control Mix Block ................................................. 93 Table 5.8: Error Analysis for the Control Mix Slab ................................................... 93

Table 5.9: Error Analysis for the 35% FA-A Mix Block ........................................... 93 Table 5.10: Error Analysis for the 50% FA-A Mix Block ......................................... 94 Table 5.11: Error Analysis for the 50% FA-A Mix Slab ............................................ 94

Table 5.12: Error Analysis for the 35% FA-C Mix Block .......................................... 94 Table 5.13: Sum of squares of prediction error by 14-day setting time method, 14-day

maturity method & 28-day maturity method vs Match cured strength ....................... 96

x

List of Figures

Figure 2.1: Temperature vs Time history of Concrete calculated according to

Equation 1 (Carino 1984)............................................................................................ 12

Figure 4.1: Arrhenius Plot for Control Mixture based on 14 days data ...................... 30 Figure 4.2: Arrhenius Plot for 35% FA-A Mix Based on 14 days data ...................... 30 Figure 4.3: Arrhenius Plot for 50% FA-A Mix Based on 14 days data ...................... 31 Figure 4.4: Arrhenius Plot for 35% FA-C Mix Based on 14 days data ...................... 31 Figure 4.5: Strength vs Equivalent Age @ 23oC for Control Mix Based on 14 days

data .............................................................................................................................. 35 Figure 4.6: Strength vs Equivalent Age @ 23oC for 35% FA-A Mix Based on 14 days

data .............................................................................................................................. 36 Figure 4.7: Strength vs Equivalent Age @ 23oC for 50% FA-A Mix Based on 14 days

data .............................................................................................................................. 37 Figure 4.8: Strength vs Equivalent Age @ 23oC for 35% FA-C Based on 14 days data

..................................................................................................................................... 38 Figure 4.9: Strength vs Equivalent Age @ 23oC for Control Mix based on Setting

Time Method ............................................................................................................... 44 Figure 4.10: Strength vs Equivalent Age @ 23oC for 35% FA-A Mix based on Setting

Time Method ............................................................................................................... 45

Figure 4.11: Strength vs Equivalent Age @ 23oC for 50% FA-A Mix based on Setting

Time Method ............................................................................................................... 46

Figure 4.12: Strength vs Equivalent Age @ 23oC for 35% FA-A Mix based on Setting

Time Method ............................................................................................................... 47 Figure 4.13: Compressive Strength vs Pullout Force for Control Mix ....................... 60

Figure 4.14:Compressive Strength vs Pullout Force for 35% FA-A Mix .................. 61

Figure 4.15: Compressive Strength vs Pullout Force for 50% FA-A Mix ................. 62 Figure 4.16: Compressive Strength vs Pullout Force for 35% FA-C Mix.................. 63 Figure 5.1: Strength Comparison for Control Mix Block by different methods ........ 87

Figure 5.2: Strength Comparison for Control Mix Slab by different methods ........... 88 Figure 5.3: Strength Comparison for 35% FA-A Block by different methods ........... 89

Figure 5.4: Strength Comparison for 50% FA-A Block by different methods ........... 90 Figure 5.5: Strength Comparison for 50% FA-A Slab by different methods ............. 91

Figure 5.6: Strength Comparison for 35% FA-C Block by different methods ........... 92

1

Chapter 1 Introduction

1.1 Research Outline

An important decision in the construction industry today revolves

around the in-place strength gain predictions and the timely removal

of formwork. Traditionally, ASTM C39 is used to estimate the

strength of concrete at the required age. Concrete samples are

subjected to same curing conditions as expected on site and samples

are tested at specific ages to determine compressive strength. The

method is simple to execute and very reliable. However, in

implementation of ASTM C39, ‘mass effect’ of concrete is often

overlooked which may provide higher in place strengths at early ages.

In the case of mass concrete structures, a large amount of heat is

generated due to exothermic reaction between the cement and water.

As a result, internal temperature of concrete is significantly higher

and concrete cures faster and thus develops strength earlier.

The objective of this research was to develop strength prediction

models which can estimate strength at early ages by minimizing

concrete testing. These approach should eventually capture the mass

concrete effects more accurately than the standard curing described in

ASTM C39.

2

Another objective of the research was to study the strength

development in High Volume Fly Ash (HVFA) concrete mixtures.

Fly Ash is a byproduct of coal used by the electric industry and is

known to have cementitious properties. About 131 million tons of fly

ash is produced annually in the United States and disposal of fly ash

poses an environmental challenge. Because of its large scale

availability, low cost and cementitious properties, and potential

improvement in concrete properties, Fly Ash became popular in the

construction industry as a replacement for cement. Small replacement

of cement by fly ash (about 5%-10%) doesn’t impact the hydration

reaction significantly but large scale replacement (about 20%-50%) is

known to slow down the hydration reaction and hence strength

develops slowly. The construction industry is averse to using HVFA

concrete due to this reason. However, the slow development is also

partly due to fact that often the mass effect is neglected.

1.2 Research Methodology

The aim was to develop strength prediction models for conventional

and HVFA mixes. The HVFA concrete mixes were developed based

on the conventional (control) mixture and using varying amounts of

two types of fly ash: Type F and Type C (these are explained further

3

in Chapter 3). In all, 4 mixes were studied (one conventional mix and

three HVFA mix) and their corresponding prediction models were

developed by using alternative methods capturing the hydration

effects. The need for considering two types of fly ash is to compare

the effects of chemical composition of fly ash on the rate of strength

development.

This research is further extension of Dr. Sushant Upadhyaya’s

research titled “Early Age Strength Prediction for High Volume Fly

Ash Concrete using Maturity Modeling.” Dr. Upadhyaya’s work was

based developing prediction models using maturity modeling based

on 28-day strength data. The rate of strength development in concrete

slows down considerably from the 14th to the 28th day. Thus, a

prediction model based on the 14-day data is may provide accurate

strength prediction values and eventually require less testing. Thus,

the focus of this research was to develop 14-day based prediction

models and then compare the results with those obtained from

alternative methods and the the 28-day Maturity approach suggested

by ASTM 1074. The research also included two alternative methods

of developing 14-day based prediction models, one based on the

4

setting time of concrete and a second one considering the constant Suc

approach (i.e., all mixtures have the same ultimate strength). Chapter

4 describes in detail the alternative modeling approaches adopted in

this research in order to develop the strength prediction models. The

major tasks in this research were as follows:

1.2.1 Activation Energy

The first step in developing prediction models for different mixes is

to determine the activation energy of the corresponding mixes.

ASTM C1074 gives the procedure for determining the activation

energy by the maturity method. As per Arrhenius, Activation energy

is defined as the minimum amount of energy required to start a

chemical reaction between potential reactants. In the context of this

research, activation energy is the energy required to start the

hydration reaction. The activation energy was calculated using three

different approaches: Variable Su Approach, Setting Time Approach

and Constant Suc Approach. These three approaches correspond to the

three different ways in which the predictive maturity models were

developed. The procedure for finding the activation energy using

these approaches is included in detail in Chapter 4.

5

1.2.2 Maturity Modeling

As mentioned earlier, models were developed for four type of

concrete mixes. Concrete testing was carried out at the previous

research phase (Upadhyaya, 2008) where concrete cylinders were

subjected to standard curing in the laboratory as per ASTM C39.

Cylinders were made corresponding to each of the four mixes. The

cylinders were then tested at the ages of 1,2,4,7, 14 & 28 day to

determine the corresponding compressive strength as per ASTM C39.

At each testing age, an average of three test cylinders was reported as

compressive strength for that age. Temperature sensors were placed

in two of the concrete cylinders to monitor temperature profile during

curing. The raw data for strength at respective ages for all the mix

was available through Dr. Upadhyaya’s earlier work. Using the

activation energy, the actual age at 1,2,4,7 & 14 day were converted

to equivalent age @ 230 C. Chapter 2 explains in detail such

procedure.

1.2.3 Match Curing

As mentioned earlier, large concrete structures (mass concrete

effects) may experience higher internal temperatures due to the

6

exothermic reaction between the cement and water. Inability to take

this into account will lead to under-prediction of strength. To account

for such effects the match curing approach was adopted. In match

curing, the cylinders to be tested for strength are subjected to the

same temperature profile as the large concrete structure on site. This

is achieved by inserting temperature sensors inside the concrete, on

site, and at specific locations within the mass structure. The

temperature profile of concrete can then be monitored and used to

cure the concrete cylinders. The higher temperature causes faster

curing and concrete gains strength at a faster rate. The strength

achieved from match curing is eventually representing most

accurately the on-site concrete strength. Thus, the accuracy of the

strength prediction models can be judged on how well the predicted

values compare with those from the match cured strength values. Due

to testing logistics the match curing method adopted in the previous

phase of this research was limited to 2, 4 and 7 days (Upadhyaya,

2008). Thus since match curing strength was not available at 28

days, a part of this research was dedicated to predict the match cured

strength at 28 days. Based on the experimental data and past attempts

suggested in the literature (discussed in Chapter 2), a basic structure

7

for the model development process was identified and the best-fit

equation was used. This allowed to predict the 28-day strength. The

procedure, analysis and discussion regarding match curing is further

addressed in Chapter 4.

1.2.4 Field Curing

Field curing testing was also included in the previous phase of the

research and the data were used for the analysis of this work. The

field cured cylinders were tested for strength at 2, 4 & 7 days. For

each age, the average of 3 cylinders was recorded as compressive

strength. The field curing strength data are discussed in in Chapter 4.

1.2.5 Pullout Tests

ASTM C900 describes the procedure for determining the pullout

strength of concrete. A metal is inserted in concrete and this test

determines the force (known as pullout force) required to remove the

metal insert from the concrete structure. This is a non-destructive test.

The idea is to determine the pullout force and use the pullout force vs

compressive strength to relate the two. Usually, the manufacturers of

pullout testing apparatus provide the correlations between pullout

8

strength and compressive strength. As per ACI 228.1R-03, this

correlation is of the following form: C = a x Pb

Where: C = Compressive strength MPa (psi),

P = Pullout force (kN)

a , b = Regression constants (MPa, psi)

The pullout test data was available from the past research phase and

to have better strength predictions, the correlation between pullout

force and compressive strength were developed for these specific

concrete mixtures. This analysis is discussed in Chapter 4.

1.3 Organization of Thesis

Chapter 2 provides the background for maturity modeling and its

development. The methods for calculating the activation energy are

discussed in that chapter as well.

Chapter 3 provides information on raw materials used in this

research. The chemical composition of different types of fly ash, mix

design proportion for the conventional and HVFA concrete mixtures

9

along with the source of raw materials are summarized in detail in

this chapter.

Chapter 4 includes the modeling analysis and results. It describes the

different approaches that were adopted to develop the prediction

models. The procedure for calculating activation energy, maturity

modeling, match curing, field curing and pullout testing are explained

in detail along with the results.

Chapter 5. This chapter describes the results from alternative methods

of maturity modeling. The predicted values of strength from the

prediction models developed in this research are compared with the

models developed in the previous phase of the study (Upadhyaya,

2008) to determine the best approach for each mixture.

Chapter 6 provides the conclusions. The findings of the modeling

and their accuracy is summarized in this chapter.

10

Chapter 2 Background

2.1 Development of Maturity Functions

Concrete gains strength over a period of time. The rate of strength

gain of concrete can be accelerated, among other, by curing the

concrete at higher temperature. The rate of strength gain can be

accelerated by curing at higher temperature, but such effects are not

expected on the ultimate strength. Thus, the rate of strength gain in

concrete can be controlled by increasing curing temperature and

increasing curing time (or age), among other means like

proportioning and use of chemical admixtures. The effects of

temperature and time on concrete can be explained by the term

‘Maturity Function’.

As explained in ASTM C1074, maturity function is a mathematical

expression which summarizes the time-temperature history of

concrete (or cementitious mixture) during the curing period to

calculate an index known as maturity index. The main objective of

this function is to explain the combined effects of time and

temperature on strength development at elevated curing temperatures.

11

The Nurse-Saul equation is a popular maturity equation named after

the works of Nurse (1949) and Saul (1951) and is defined in ASTM

C1074.

The Nurse-Saul Equation is given by

Equation 1

Where, M = Nurse-Saul maturity index at age t (°C • hours),

Ta = average concrete temperature during the ∆t (°C),

To = datum temperature (°C), and

∆t = time interval (hours).

Thus, according to Nurse-Saul equation, the product of temperature

and time is good representation of maturity of concrete. It is

important to note that the process of strength development starts

above a temperature known as datum temperature. The minimum

temperature above which concrete begins to develop strength is

known as Datum Temperature.

The basic idea is that concrete of same mix with same maturity index

(product of temperature and time) will have same strength

irrespective of the curing history. For example, concrete cured at 25

12

°C for 10 days will have same strength as concrete cured at 50 °C for

5 days, if they are of the same mix. This is because both have the

same maturity index of 250 °C • Day. The Nurse-Saul equation

represents one of earliest works in developing maturity functions.

Figure 2.1: Temperature vs Time history of Concrete calculated according to

Equation 1 (Carino 1984)

According to Nurse-Saul equation, the rate of strength gain is a linear

function of time. However, this is not true. When cement (or

cementitious material) and water are mixed, there is time delay before

the process of strength development begins. In other words, the

process of strength development is not instantaneous. This time delay

period is called Induction period (Carino and Lew, 2011). After the

13

induction period is the acceleratory period: the period where strength

develops rapidly (Carino and Lew, 2011). The strength develops

rapidly for initial days and beyond about 28-days, the rate of strength

development becomes slower and slower. Due to this, a linear

approximation is not a very accurate measure of maturity and there

was a need to come up with alternatives for the widely accepted and

popular Nurse-Saul equation.

The age of concrete is relative to the curing temperature. At higher

temperatures, concrete cures faster. The standard curing temperature

as defined in ASTM C39 is 23°C. Thus, if the curing is carried out at

any temperature other than 23°C, the practice is to represent age of

concrete in terms of equivalent age @ 23°C.

The equivalent age @ 23°C is defined as the age of concrete @ 23°C

for which the concrete would have had the same maturity had it been

cured at 23°C. The equivalent age of concrete is mixture specific i.e.

the equivalent age will be different for different mixes. If the rate of

hydration reaction is faster, the concrete matures quickly and hence,

the equivalent age @ 23°C will be low. Conversely, if the rate of

14

reaction is slower, it takes time for concrete to mature and gain

strength and therefore, the equivalent age @ 23°C will be higher. The

rate at which the reaction proceeds, therefore, becomes an important

factor in maturity modeling. The rate of reaction is decided by the

‘Activation Energy’ of the mix.

Activation Energy is defined as the minimum energy required for

reaction to start and proceed. Arrhenius developed an equation, now

known as the Arrhenius Equation, which captured the temperature

dependence of reaction rates through the concepts of activation

energy. Freiesleben, Hansen and Pedersen (1977) developed a new

maturity function based on the ideas of Arrhenius equation.

This equivalent age equation is as follows:

Equation 2

te = the equivalent age at the reference temperature,

E = apparent activation energy, J/mol,

R = universal gas constant, 8.314 J/mol-K,

T = average absolute temperature of the concrete during curing

period, Δt

15

Tr = absolute reference temperature, Kelvin

In this research, the above maturity function is used to compute the

equivalent ages.

2.2 Strength-Equivalent Age Relationship

The aim of this research is to develop models which can predict the

strength of concrete at required age with fair accuracy. To be able to

do this, it is essential to know the variation of strength over a period

of time. It is already well known that concrete begins to develop

strength after the induction period and develops strength rapidly

during the early ages and rate of strength development slows down

with time. However, concrete continues to gain strength throughout

the course of its life. The challenge now is to capture this behavior

through a mathematical model.

Carino (1984) developed a hyperbolic equation which could capture

this behavior provided the curing was carried out under isothermal

conditions up to equivalent ages of 28-days at standard curing

temperature of 23 °C.

16

This equation is as follows:

Equation 3

Where, St = strength at age

Su = limiting strength,

k = rate constant, 1/day, and

t0 = age at start of strength development.

In the above hyperbolic equation, the strength is assumed to develop

after the final setting time of the mixture. The values of setting time

for all the mixtures are given in table 4.1.

Equation 3 is used in this research for developing maturity models to

predict strength of concrete at required ages.

The term St represents the strength of concrete at age‘t’ days.

Following the guidelines given in ASTM C39, the strength of

concrete can be found out at the required age. Thus, we can find out

the term St and the corresponding age of the mix as well. Once this

data is obtained, the best fit curve is applied to the data and the from

17

the equation of this best fir curve, the rate constant ‘k’ and limiting

strength ‘Su’ can be found out.

In this research, Matlab was used to develop the strength – equivalent

age @ 23°C relationship using equation 3.

18

Chapter 3 Materials & Concrete Mix Design

This research follows the initial work carried out in a previous

experimentation (Upadhyaya 2008) titled ‘Early Age Strength

Prediction for HVFA Concrete Using Maturity Modeling’.

3.1 Raw Materials

A no.57 crushed limestone coarse aggregate and natural sand

conforming to ASTM C33 was used to prepare the concrete mixtures

and samples for testing. Table 3.1 lists the gradation properties for the

aggregates used. The gradation for the coarse and fine aggregates was

carried out as per the provisions of ASTM C136. The ASTM C136

provides the procedure for sieve analysis of coarse and fine

aggregates.

19

Table 3.1: Gradation of Coarse and Fine Aggregates

Percent Passing

Sieve Sizes Coarse Aggregate Fine Aggregate

No 57 -

1 ½ 100 0

1 100 0

¾ 92 0

½ 49 0

3/8 28 100

No 4 5 99

No 8 1 84

No 16 0 70

No 30 0 52

No 50 0 20

No 100 0 3

No 200 1 -

Apart from gradation analysis, the aggregates were also tested for

some specific properties like absorption, specific gravity etc. Table

3.2 provides the properties of coarse and fine aggregates used in this

research.

20

Table 3.2: Properties of Coarse and Fine Aggregates

Properties Coarse

Aggregate

Fine

Aggregate

Fineness Modulus - 2.73

Specific Gravity (SSD) 2.84 2.59

Absorption,% 0.30 1.30

Dry rodded unit weight,

lb/ft3

105.90 N/A

Type I Portland Cement conforming to ASTM C150 was used for

preparing the concrete mixes. Two types of fly ash were used

conforming to specification of ASTM C618:

1. Class F fly ash having CaO content of 1.0 %. This fly ash

wasidentified as FA-A henceforth throughout the course of this

research.

2. Class C fly ash having CaO content of 23.44 %. This fly ash was

identified as FA-C henceforth throughout the course of this research.

21

Apart from this, Polycarboxylate based Type F High Range Water

Reducer (HRWR) conforming to specifications of ASTM

C494/C494M was used.

The fly ash and the HRWR admixtures were procured from the

following sources:

i) FA-A was procured from STI, Baltimore, MD,

ii) FA-C was procured from Boral Material Technologies Inc.,

iii) HRWR admixture was supplied by Sika Corporation.

3.2 Mixture Proportion

For ease of testing and convenience, testing the actual concrete

specimen is not preferred. Whenever possible, corresponding mortar

mixes are prepared such that they preserve the integrity of the actual

concrete mix. This can be achieved by proportioning the mortar

mixes such that the fine aggregate-to-cementitious materials ratio (by

mass) is the same as coarse aggregates-to-cementitious materials

ratio. This is consistent with the recommendations of ASTM C1074

Annex A1. A major task in this research involved determining the

activation energy. The activation energy was determined using the

22

mortar mixes by proportioning them as described above. The actual

mix design proportions are tabulated below.

Table 3.3: Mix Design Specifications for Mortar Mixes

Item Control

Mix

35 % FA-A

Mix

50 % FA-A

Mix

35 % FA-C

Mix

Cement (grams) 1876.00 1199.00 1101.00 1357.00

Fly Ash (grams) 0.00 710.00 1066.00 740.00

Fine Aggregate (grams) 7136.00 7087.00 7036.00 7250.00

Water (grams) 1052.00 960.00 848.00 889.00

HRWR Admixture (ml) 62.90 200.67 212.64 152.75

w/cm 0.56 0.51 0.39 0.42

The next part of research focused on developing the prediction

models for different concrete mixtures. The process of developing

these models is explained in detail in Chapter 4. As a part of this

process, standard concrete cylinders 10.2cm X 20.3cm (4in by 8in)

were required to be made for all the mixtures given in table 3.3 for

purpose of compression testing as per ASTM C39. The mix design

proportions for making these cylinders is given in table 3.4.

23

Table 3.4: Mix Design Proportions for Concrete Cylinders

Item Control

Mix

35 % FA-A

Mix

50 % FA-A

Mix

35 % FA-C

Mix

Cement, kg/m3 302.60 196.40 182.70 215.40

Fly Ash, kg/m3 0.00 116.30 176.80 117.50

Coarse Aggregate, kg/m3 1151.00 1160.40 1167.00 1151.00

Fine Aggregate, kg/m3 770.10 752.30 769.50 783.70

Water, kg/m3 169.70 157.20 140.60 141.20

HRWR Admixtures,

ml/45kg

62.90 200.70 140.60 152.70

w/cm 0.56 0.50 0.39 0.42

24

Chapter 4 Experimental Work, Analysis and Results

The aim of this research is to develop maturity based prediction

models which can predict the strength of concrete fairly accurately

for required ages. To this effect, different approaches were tried. The

determination of activation energy of the four different concrete

mixes under consideration was one of the major tasks of this research.

4.1 Determination of Activation Energy

4.1.1 Activation Energy by ASTM C1074

ASTM C1074 describes the procedure for estimating the concrete

strength by maturity method. The process of determining activation

energy can be considered as a two-step process. Firstly, the rate

constant for the reaction is determined and from the rate constant, the

activation energy can be determined. Initially, mortar mixes

representative of respective PCC and HVFA mixes are prepared.

Annex A1 of ASTM C1074 clearly states that values of activation

energy obtained by analyzing mortar mixes are applicable to

corresponding concrete mixes. Once the cubes of mortar mixes are

cast, compressive strength test is performed on them as per ASTM

25

C39 at various ages. Thus, for a specific mix, the aim is to achieve a

set of values of strength of that mix at various ages under different

curing conditions. Once this data is obtained, Equation 3 as described

in chapter 2 is used to find the rate constant. This equation is as

follows:

In the above equation, ‘k’ represents the rate constant. The values of

St were recorded for t = 1,2,4,7 and 14 days. This is because the

primary aim of this research is to develop maturity models based on

14-day data and hence, even though available, the data beyond 14-

days was included only for comparison The values of St are plotted

against respective values of age t and best fit curve is applied to this

set of data. This best fit curve then provides the value of rate

constant.

ASTM C1074 recommends preparing mortar cubes of size 5.08 cm (2

in) for determination of activation energy. For different mixes, the

mortar cubes were prepared based on mix design proportions given in

26

table 3.3. The cubes were cured at four different curing temperatures

of 7.5 0C (45 0F), 210C (70 0F), 380C (100 0F) and 490C (120 0F).

There were 4 different mixes and mortar cubes representing each mix

were prepared and cubes of each mix were cured at four different

temperatures as given above. For each mix at each testing age, three

5.08 cm (2 in) mortar cubes were tested and average of the three was

recorded as the compressive strength at that age.

The strength vs age data for the different mixes can be obtained from

previous phase of research (Upadhyaya, 2008). As can be seen in

Equation 3, the setting time for different mixes are required to

determine the rate constant. The setting time test was performed on

different mixes as per ASTM C403 to achieve this data. This data is

tabulated below.

27

Table 4.1: Setting Time for Mortar Mixtures (ASTM C403)

Mixture Tc=7.5 oC

(45 oF)

Tc=21 oC

(70 oF)

Tc=38 oC

(100 oF)

Tc=49 oC

(120 oF)

Initial

(hrs)

Final

(hrs)

Initial

(hrs)

Final

(hrs)

Initial

(hrs)

Final

(hrs)

Initial

(hrs)

Final

(hrs)

Control

Mix

7.80 15.00 4.70 8.50 2.90 4.10 1.80 2.50

35% FA-A

Mix

10.20 17.10 6.60 13.20 5.00 7.10 3.00 4.20

50% FA-A

Mix

10.90 19.90 7.30 14.10 5.70 8.40 3.30 5.00

35% FA-C

Mix

8.00 16.00 5.70 11.40 4.10 5.90 2.40 3.50

Table 4.1 provides the setting times for the different mortar mixes

under various curing conditions. Using the data given in this table and

strength data from previous phase of research, equation 3 as given

above can be used to find the rate constant k.

In the previous work (Upadhyaya, 2008) the rate constants for the

mixes were based on 28-day strength data while the values

determined in this research were based on 14-day strength data. Table

4.2 provides the values of rate constant based on 14-day while also

comparing the values based on 28-day strength data.

28

Table 4.2: 28-day vs 14-day k value comparison

Where kT-28 =kT value based on 28-day data

kT-14 =kT value based on 14-day data

The kT-28 values given in table above are from the previous study

(Upadhyaya, 2008) while the kT-14 were computed as a part of this

research. There was a wide variation in error when predictions were

based on 28-day model and for this reason an approach based on 14-

day model was suggested. From Table 4.2, it can be seen that every

Mixture Trial 7.5oC (45 oF)

kt (day-1)

21oC (70 oF)

kt (day-1)

38oC (100 oF)

kt (day-1)

49 oC (120oF)

kt (day-1)

kT-28 kT-14 kT-28 kT-14 kT-28 kT-14 kT-28 kT-14

Control

Mix

1 0.240 0.315 0.636 0.884 1.539 1.880 2.450 2.536

2 0.203 0.246 0.648 0.809 - - 1.973 2.134

35% FA-A

Mix

1 0.156 0.173 0.410 0.737 0.457 0.961 0.404 1.316

2 0.161 0.168 0.310 0.359 - - 0.542 0.619

3 - - 0.290 0.373 - - 0.309 0.441

50% FA-A

Mix

1 0.085 0.085 0.175 0.235 0.441 0.738 0.677 0.827

2 0.096 0.103 0.289 0.387 - - 0.772 0.856

3 - - 0.133 0.476 - - 0.666 0.592

35% FA-C

Mix

1 0.056 0.049 0.198 0.254 0.138 0.170 0.335 0.330

2 - - 0.194 0.251 - - 0.335 0.342

3 - - 0.013 0.120 - - 0.039 0.184

29

14-day rate constant value is higher than the corresponding 28-day

based rate constant value.

Once the values of rate constant have been determined, Annex A1 in

ASTM C1074 provides further directions to compute activation

energy. The natural logarithms of rate constant is calculated along

with absolute temperature of curing (kelvin = Celsius + 273). A plot

of natural logarithm of k-values is plotted against the reciprocal of

absolute temperature. These plots are known as Arrhenius plots. A

best fit line is then determined for these plots of data. The negative

slope of this line is the activation energy divided by the universal gas

constant R. Thus, activation energy can be determined by multiplying

the negative slope of best fit line by Universal Gas constant R.

The Arrhenius plots for all the four mixes are shown below. Since

multiple trials were prepared for each mix, the Activation Energy

values were determined by taking the combined data from all trials

into consideration and is denoted by ‘combined AE’ in table 4.3.

30

Figure 4.1: Arrhenius Plot for Control Mixture based on 14 days data

Activation Energy of Control Mix = 38067.55 J/Mol

Figure 4.2: Arrhenius Plot for 35% FA-A Mix Based on 14 days data

Activation Energy of 35% FA-A Mix = 29706.04 J/Mol

31

Figure 4.3: Arrhenius Plot for 50% FA-A Mix Based on 14 days data

Activation Energy of 50% FA-A Mix = 34889.02 J/Mol

Figure 4.4: Arrhenius Plot for 35% FA-C Mix Based on 14 days data

Activation Energy of 35% FA-C Mix = 24627.51 J/Mol

The values of activation energy are shown in Table 4.3

32

Table 4.3: Activation Energy (ASTM C1074) Based on 14 days data

Mixture AE (Trial 1) (J/Mol)

AE (Trial 2) (J/Mol)

AE (Trial 3) (J/Mol)

Combined AE (J/Mol)

Control Mix

37619.90 37672.75 NA 38067.55

35% FA-A Mix

33996.27 22438.81 NA 29706.04

50% FA- A Mix

43194.95 36346.59 6093.69 34889.02

35% FA-C Mix

27887.57 8712.37 NA 24627.51

Once the activation energy was determined, compressive strength test

was performed on standard cylinders of size 10.2 X 20.3 cm (4in by

8in) as per ASTM C39. At each testing age, an average of three

testing cylinders was reported as compressive strength. Temperature

sensors were embedded in two cylinders to monitor the temperature

profile during curing. Using the activation energy computed above

for each mix, the age at which compressive strength test was

performed on these standard cylinders were converted into equivalent

age @ 23 oC. From the strength as determined from ASTM C39 and

the corresponding equivalent age @ 23 oC, equation 3 was used to

develop 14-day based maturity model for each of the mix. The results

for conversion of age into equivalent age @ 23 oC and the

corresponding compressive strength for all the mixes are given in the

33

following tables. Furthermore temperature sensors were embedded in

cylinder to maintain their temperature same as concrete cubes

prepared for pullout testing. This is discussed in detail in section 4.

Table 4.4: Compressive Strength and Equivalent Age for Control Mix

Based on 14 days data

Age Equivalent Age @ 23 oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.90 7.10

2 1.80 11.80

4 3.60 16.90

7 6.30 18.60

14 12.60 23.90

Table 4.5: Compressive Strength and Equivalent Age for 35% FA-A Mix

Based on 14 days data

Age Equivalent Age @ 23 oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.92 4.80

2 1.84 7.10

4 3.68 9.70

7 6.45 12.60

14 12.90 18.00

34

Table 4.6: Compressive Strength and Equivalent Age for 50% FA-A Mix

Based on 14 days data

Age Equivalent Age @ 23 oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.91 7.20

2 1.82 11.50

4 3.63 16.40

7 6.36 19.50

14 12.71 25.30

Table 4.7: Compressive Strength and Equivalent Age for 35% FA-C Mix

Based on 14 days data

Age Equivalent Age @ 23 oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.93 5.60

2 1.87 12.30

4 3.74 19.50

7 6.54 24.20

14 13.08 38.30

From the equivalent age @ 23 oC and the compressive strength data

provided in the above tables, prediction models were developed for

each mix. These prediction models are shown in Figures 4.5 to 4.8.

35

0 2 4 6 8 10 12 14 160

5

10

15

20

25

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.5: Strength vs Equivalent Age @ 23oC for Control Mix Based on 14 days

data

The equation of the parabolic function of Figure 4.5 is the prediction

model for the control mix based on the activation energy given in

Table 4.3. This prediction model is given by

St = 10.84*t /(1+0.3861*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

36

0 2 4 6 8 10 12 140

2

4

6

8

10

12

14

16

18

20

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.6: Strength vs Equivalent Age @ 23oC for 35% FA-A Mix Based on 14 days

data

The equation of the parabolic function of figure 4.6 is the prediction

model for 35% FA-A mix based on activation energy given in Table

4.3. This prediction model is given by

St = 4.89*t /(1+0.2048*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

37

0 2 4 6 8 10 12 14 160

5

10

15

20

25

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.7: Strength vs Equivalent Age @ 23oC for 50% FA-A Mix Based on 14 days

data

The equation of the parabolic function of figure 4.7 is the prediction

model for 50% FA-A mix based on activation energy given in Table

4.3. This prediction model is given by

St = 9.67*t /(1+0.3135*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

38

0 2 4 6 8 10 12 14 160

5

10

15

20

25

30

35

40

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.8: Strength vs Equivalent Age @ 23oC for 35% FA-C Based on 14 days data

The parabolic function of the figure 4.8 is the prediction model for

35% FA-C mix based on activation energy given in Table 4.3. This

prediction model is given by

St = 9.761*t /(1+0.2603*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

39

4.1.2 Activation Energy by Setting Time Method

The setting of mortar or concrete is the gradual transition from liquid

to solid state. ASTM C403 was performed on mortar mixtures to

determine the setting time. These results are tabulated in table 4.1.

Setting time is defined as the time interval between initial setting and

final setting. Mathematically, this can be represented as follows:

Equation 4

Where: ST = Setting Time (mins)

FS = Final Setting Time (mins)

IS = Initial Setting Time (mins)

The initial and final setting time are determined using the Vicat’s test.

A 1mm diameter needle is allowed to penetrate freshly prepared

cement paste for 30 seconds and the penetration is noted. The initial

setting time is the time required for a penetration of 25mm to take

place while the final setting time is when the needle does not

penetrate into the paste. The hydration reaction starts as soon as

cement comes in contact with water. As soon as the hydration

40

reaction starts, the hydration temperature continues to increase until

final setting time of concrete is reached. The final setting time relates

to point at which stresses and stiffness begins to develop in freshly

prepared concrete.

When concrete is cured at higher temperature, the hydration reaction

will proceed at a faster rate thereby reducing the initial and final

setting times. Conversely, concrete cured at lower temperature will

increase the setting time of concrete. Thus, hydration reaction is the

inverse of setting time. Mathematically, this can be represented as:

Equation 5

Where: Kt = Hydration Rate

Once the hydration constant is computed, the Activation Energy is

determined in the same way as described in ASTM C1074 Annex A1.

The activation energy computed by this approach is provided in the

Table 4.8.

41

Table 4.8: Activation Energy by Setting Time Approach

Mixture Activation Energy (J/Mol)

Control Mix

43328.04

35% FA-A Mix

40809.05

50% FA-A Mix

31373.57

35% FA-C Mix

37740.38

Based on the values of activation energy given in table 4.8, prediction

models were developed. All the data to develop these models are

shown in Table 4.9-4.12.

Table 4.9: Compressive Strength and Equivalent Age for Control Mix

based on Setting Time Method

Age Equivalent Age @ 23oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.89 7.10

2 1.77 11.80

4 3.55 16.90

7 6.21 18.60

14 12.42 23.90

42

Table 4.10: Compressive Strength and Equivalent age for 35% FA-A mix

based on Setting Time Approach

Age Equivalent Age @ 23oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.91 4.80

2 1.82 7.10

4 3.64 9.70

7 6.38 12.60

14 12.75 18.00

Table 4.11: Compressive Strength and Equivalent Age for 50% FA-A

Mix based on Setting Time Method

Age Equivalent Age @ 23oC

(Days)

Strength

(MPa)

0 0.00 0.00

1 0.92 7.20

2 1.83 11.50

4 3.67 16.40

7 6.42 19.50

14 12.84 25.30

43

Table 4.12: Compressive Strength and Equivalent Age for 35% FA-C

Mix based on Setting Time Method

Age Equivalent Age @ 23oC

(Days)

Strength

(Mpa)

0 0.00 0.00

1 0.90 5.60

2 1.80 12.30

4 3.60 19.50

7 6.31 24.20

14 12.61 28.30

From the equivalent age @ 23 oC and the compressive strength data

provided in Tables 4.9 to 4.12, prediction models were developed for

each mix. These prediction models are presented next.

44

0 2 4 6 8 10 12 140

5

10

15

20

25

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.9: Strength vs Equivalent Age @ 23oC for Control Mix based on Setting

Time Method

The equation of the curve above is the prediction model for Control

mix based on activation energy found by setting time method. This

prediction model is given by

St = 11*t / (1+0.3917*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

45

0 2 4 6 8 10 12 14 160

2

4

6

8

10

12

14

16

18

20

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.10: Strength vs Equivalent Age @ 23oC for 35% FA-A Mix based on Setting

Time Method

The equation of the curve above is the prediction model for 35% FA-

A based on activation energy found by setting time method. This

prediction model is given by

St = 4.944*t /(1+0.2070*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

46

0 5 10 150

5

10

15

20

25

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.11: Strength vs Equivalent Age @ 23oC for 50% FA-A Mix based on Setting

Time Method

The equation of the curve above is the prediction model for 50% FA-

A mix based on activation energy found by setting time method. This

prediction model is given by

St = 9.576*t / (1+0.3104*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

47

0 2 4 6 8 10 12 140

5

10

15

20

25

30

Equivalent Age @ 23oC (Days)

Str

ength

(M

Pa)

Figure 4.12: Strength vs Equivalent Age @ 23oC for 35% FA-A Mix based on Setting

Time Method

The equation of the curve above is the prediction model for 35% FA-

C mix based on activation energy found by setting time method. This

prediction model is given by

St = 10.12*t / (1+0.2698*t)

Where St is the compressive strength and ‘t’ is the equivalent age @

23oC.

48

4.1.3 Activation Energy by Constant Suc Approach

In most situations, concrete mixtures are designed to have certain

amount of ultimate compressive strength. Hence, the mix design

process of concrete focuses on attaining a required amount of

compressive strength known as the ultimate strength. In section 4.1.1,

the best fit trend line gives variable value of ultimate strength

depending on the curing conditions. Ideally, there should be one

limiting strength value (ultimate strength) for concrete mixtures

irrespective of the temperature at which they were cured. In this

approach, the activation energy is calculated as per ASTM C1074 just

as in section 4.1.1. However, instead of variable Su achieved due to

best fit trend line, the limiting strength was kept constant (Suc) and

then the best fit line provided the values of rate constant. The values

of limiting strength and rate constant are given in Table 4.13

It must be noted that this approach is similar to the one employed in

section 4.1.1. However, there is a conceptual difference between the

two approaches. While the approach in section 4.1.1 will yield

multiple values of limiting strength based on function of the curing

49

conditions, this approach yields one value of limiting strength

irrespective of curing conditions.

Table 4.13: Rate Constants using Constant Suc Approach

Mixture Trial Suc

(MPa)

7.5oC(45 oF)

kt (day-1)

21oC (70 oF)

kt (day-1)

38oC(100 oF)

kt (day-1)

49 oC(120oF)

kt (day-1)

Control

Mix

1 31.30 0.207 0.719 1.492 1.760

2 29.20 0.261 0.625 - 1.419

35% FA-A

Mix

1 28.30 0.047 0.228 0.677 0.923

2 29.80 0.036 0.151 - 0.939

3 28.10 0.042 0.242 - 0.732

50% FA-A

Mix

1 53.20 0.042 0.140 0.499 0.718

2 46.60 0.073 0.187 - 0.968

3 50.00 - 0.132 - 0.720

4 28.10 - 0.501 - 1.849

35% FA-C

Mix

1 42.70 0.039 0.135 0.163 0.696

2 41.00 - 0.111 - 0.467

3 44.50 - 0.099 - 0.393

From the rate constants provided in Table 4.13 the value of activation

energy was determined based on all the trials. This value of activation

energy is denoted by ‘Combined AE’ in Table 4.14.

50

Table 4.14: Activation Energy by Constant Suc Approach

Mixture AE (Trial 1) ( J/Mol)

AE (Trial 2) ( J/Mol)

AE (Trial 3) ( J/Mol)

AE (Trial 4) ( J/Mol)

Combined AE

( J/Mol)

Control Mix

38666.89 29738.30 NA NA 34720.67

35% FA-A Mix

46511.35 NA NA NA 46511.35

50% FA-A Mix

52556.75 46774.40 47547.37 36381.84 49415.46

35% FA-C Mix

45363.32 40423.53 38679.25 NA 42692.49

The addition of fly ash slows the rate of hydration reaction in

concrete. This is also evident in table 4.1 where the setting time of

concrete increases with increase in fly ash content. Thus, it is

expected that increasing fly ash content will lead to lower value of

activation energy. The results of Table 4.14 that are based on the

activation energy computed by the constant Suc approach do not

represent the effect of fly ash on concrete. Hence, this approach was

judged to not represent well the hydration of these mixtures.

4.2 Actual in-place strength due to mass concrete effects

The developed prediction models provide engineers the opportunity

to predict the in place compressive strength of concrete members.

51

The degree of accuracy of these prediction models depend upon how

the predicted values from these models compare against the actual in-

place strength after considering the mass effect in concrete structures.

This in-place strength also known as match cured strength is the most

accurate estimate of in place strength of concrete members.

Temperature sensors are inserted in concrete structure at specific

locations to capture the mass effect. These sensors capture the actual

temperature profile inside the structure and transmit the data on to a

micro-controller. The micro-controller, with the help of a software

replicates the same temperature profile inside the standard concrete

cylinders (as required in ASTM C39) with the help of thermocouples.

Compression test is then performed on these cylinders at required

ages to determine the actual in place strength. The process is called

match curing simple because the temperature profile of cylinders is

matched with that of the actual structure during curing.

The challenge now was to build a structure large enough to observe

the mass effect of concrete. For detailed analysis, two types of

structures were built: a concrete block of 0.6 x 0.6 x 1.8 m (2 x 2 x 6

52

ft) and concrete slab 2.4 x 2.4 x 0.18 m (8ft x 8ft x 7 in). Due to

logistical issues, the concrete slab was developed only for the control

mix and 50% FA-A mix.

4.2.1 Block Testing

Since the research focused on analyzing 4 concrete mixes, 4 concrete

blocks were cast with each block representing each mix. It is

important to note that the concrete blocks were used to serve two

purposes: to simulate the mass effect and for performing pullout test.

The pullout test is a non-destructive test and is explained in detail in

section 4.3. In this section, the use of concrete block to simulate mass

effect will be discussed.

Temperature sensors (also known as iButtons) were used to map the

temperature profile of the concrete block. These temperature sensors

were inserted at different locations in the block. For the purpose of

maturity, the data from the sensor located at 2.54 cm from surface of

block with is relayed to micro-controller. The micro-controller

matches the temperature of the cylinders with this temperature with

the help of thermocouple. This thermocouple should be inserted at

53

mid-depth of the cylinder with a 1in cover from the edges for

accurately simulating the temperature in the cylinder. The activation

energy for the mixtures have been previously determined and with the

temperature data from iButton, the equivalent age of the block is can

be computed using equation 2 given in chapter 2. With the equivalent

age, the compressive strength is predicted from the prediction models

developed above.

The match cured strength is determined at the ages of 2, 4 and 7 days.

At each age, three cylinders are tested and the average is reported as

the compressive strength. This strength is the best estimate of in place

compressive strength and is known as match cured strength and the

value obtained from prediction model will be compared against this.

Although the 28-day match cured strength was not available, it was

required for purpose of comparisons. Equation 3 was used to predict

the match cured strength at 28 days. The existing strength values

were substituted as St and their corresponding equivalent ages @

23oC were computed from the measured temperature profile. A trend

line was fitted to this data and from this best fit tend line, the values

of limiting strength Su and rate constant ‘k’ were determined and

54

attempt was made to predict the 28-day match cured strength. The

results of match curing are presented in Table 4.15

Table 4.15: Match Cured Strength-Control Mix Block

Actual Age

(Days)

Match Cured Strength

(MPa)

0 0.00

2 19.33

4 23.85

7 26.51

28 29.84

Table 4.16: Match Cured Strength-35% FA-A Mix Block

Actual Age

(Days)

Match Cured Strength

(MPa)

0 0.00

2 12.54

4 16.63

7 19.34

28 23.09

Table 4.17: Match Cured Strength-50% FA-A Mix Block

Actual Age

(Days)

Match Cured Strength

(MPa)

0 0.00

2 14.92

4 19.48

7 22.42

28 26.41

55

Table 4.18: Match Cured Strength-35% FA-C Mix Block

Actual Age

(Days)

Match Cured Strength

(MPa)

0 0.00

2 23.68

4 30.23

7 34.29

28 39.62

4.2.2 Slab Testing

Due to logistical issues, the concrete slabs were prepared only for two

mixtures, the control mix and 50% FA-A mix. Similar to blocks, even

the slab were used for match curing as well as pullout testing. The use

of slabs for pullout testing will be discussed in section 4.3. In the

slabs, the temperature sensors located at 5.08 cm from top surface

around middle third of the slab were used for maturity calculation.

The match cured strength test was performed at 2, 4 and 7 days. The

process of predicting the match cured strength at 28 days is exactly

the same as described for concrete blocks.

56

Table 4.19: Match Cured Strength-Control Mix Slab

Actual Age

(Days)

Match Cured Strength

(MPa)

0 0.00

2 19.29

4 25.40

7 29.39

28 34.86

Table 4.20: Match Cured Strength-50% FA-A Slab

Actual Age

(Days)

Match Cured Strength

(MPa)

0 0.00

2 10.69

4 14.90

7 17.92

28 22.48

4.3 Pullout Testing

Pullout test is a non-destructive test. Similar to match curing, the

pullout test is used to determine the in place compressive strength of

concrete. The pullout test measures the force needed to extract an

embedded insert from concrete mass (Carino 2003). This force is

known as Pullout force. This pullout force is then used to determine

57

the compressive strength of concrete by using some previously

established compressive strength vs pullout force relationships.

Concrete cubes 20.32 cm (8 in) size, used previously for determining

the actual in place strength of concrete blocks, were also used to

develop the compressive strength vs pullout load relationship. To

achieve this, sufficient concrete cubes were casted such that at the

required age, one cube was used to determine the actual in place

strength while pullout test was performed on other cube. Temperature

sensors were inserted at height of 2.54 cm (1 in) from bottom surface

of the cube at the center. These sensors helped in maintaining the

temperature of cubes and the cylinders. Pullout inserts were

embedded in each of the four sides of the cube. This was done to

prevent any radial cracking. The testing was conducted at age of

1,2,4,7 and 14 days Upadhyaya 2008).

4.3.1 Compressive Strength vs Pullout Force Correlation

Compressive strength has been related to the pullout force with

exponential functions. The manufacture of pullout testing apparatus

provide such relationships for use. . However, in this research such

58

relationship was developed for the specific mixtures used. The

general form of such relation as given in ACI 228.1R-03 is:

Equation 6

Where, C = Compressive Strength (MPa)

a,b = Regression Constants, a (MPa)

P = Pullout Force (kN)

The results from pullout testing are provided below:

Table 4.21: Pullout Force and Compressive Strength for Control Mix

Age

(Days)

Pullout

Force (kN)

Standard Cured

Strength (MPa)

0 0.00 0.00

1 8.45 7.10

2 12.50 11.80

4 15.63 16.90

7 17.90 18.60

14 21.61 23.90

Table 4.22: Pullout Force and Compressive Strength for 35% FA-A Mix

Age

(Days)

Pullout

Force (kN)

Standard Cured

Strength (MPa)

0 0.00 0.00

1 7.19 4.80

2 9.40 7.10

4 10.59 9.70

7 13.41 12.60

14 18.03 18.00

59

Table 4.23: Pullout Force and Compressive Strength for 50% FA-A Mix

Age

(Days)

Pullout

Force (kN)

Standard Cured

Strength (MPa)

0 0.00 0.00

1 10.44 7.20

2 13.97 11.50

4 17.36 16.40

7 19.58 19.50

14 25.45 25.30

Table 4.24: Pullout Force and Compressive Strength for 35% FA-C Mix

Age

(Days)

Pullout

Force (kN)

Standard Cured

Strength (Mpa)

0 0.00 0.00

1 7.16 5.60

2 12.86 12.30

4 18.30 19.50

7 20.84 24.20

14 22.49 38.30

Based on the data given in tables 4.21 to 4.24 and equation 6, the

relationships between compressive strength and pullout force were

determined.

60

0 5 10 15 20 250

5

10

15

20

25

Pullout Force (kN)

Sta

ndard

Cure

d C

om

pre

ssiv

e S

trength

(M

Pa)

Figure 4.13: Compressive Strength vs Pullout Force for Control Mix

The correlation between Compressive strength and Pullout load is

given by

C = 0.5009*P1.259

61

0 2 4 6 8 10 12 14 16 18 200

2

4

6

8

10

12

14

16

18

20

Pullout Force (kN)

Sta

ndard

Cure

d C

om

pre

ssiv

e S

trength

(M

Pa)

Figure 4.14: Compressive Strength vs Pullout Force for 35% FA-A Mix

The correlation between Compressive strength and Pullout load is

given by

C = 0.3733*P1.345

62

0 5 10 15 20 250

5

10

15

20

25

Pullout Force (kN)

Sta

ndard

Cure

d C

om

pre

ssiv

e S

trength

(M

Pa)

Figure 4.15: Compressive Strength vs Pullout Force for 50% FA-A Mix

The correlation between Compressive strength and Pullout load is

given by

C = 0.3755*P1.309

63

0 5 10 15 20 250

5

10

15

20

25

30

35

40

45

Pullout Force (Mpa)

Sta

ndard

Cure

d C

om

pre

ssiv

e S

trength

(M

Pa)

Figure 4.16: Compressive Strength vs Pullout Force for 35% FA-C Mix

The correlation between Compressive strength and Pullout load is

given by

C = 0.04584*P2.121

64

4.3.2 Pullout Tests on concrete blocks

The concrete blocks of size 0.6 x 0.6 x 1.8 m that were casted for

match curing as described in section 4.2 were also used for pullout

testing. 12 Pullout inserts were embedded on each of the longer side.

Thus, there were 24 inserts in total. ASTM C900 provides guidelines

with regards to minimum distance between inserts, clear cover

between the edges and inserts etc. Following these guidelines, the

inserts were placed at 0.145 m distance center to center and 0.115 m

from the edge. The pullout inserts extended 2.54 cm into the concrete

surface. The pullout test were performed at 2, 4 and 7 days using the

“SureCure” system (Upadhyaya, 2008).

4.3.3 Pullout testing on concrete slabs

The concrete slabs that were casted for match curing of size 2.4 x

2.4x 0.18 m were also used for pullout testing. Due to logistical

issues, the slabs were casted only for control mix and 50% FA-A

mix. The inserts were embedded in slab just the same way as they

were embedded in block as described above. The pullout tests were

performed at 2, 4 and 7 days. From the pullout forces, the

compressive forces were estimated from the models developed above

65

(Upadhyaya, 2008). The results of the pullout test are provided in

Table 4.25 to 4.30.

Table 4.25: Pullout Force and estimated Compressive Strength for

Control Mix Block

Age

(Days)

Pullout Force

(kN)

Estimated Strength

(MPa)

0 0.00 0.00

2 17.18 18.00

4 18.53 19.73

7 21.53 23.84

Table 4.26: Pullout Force and Compressive Strength for Control Mix

Slab

Age

(Days)

Pullout Force

(kN)

Estimated Strength

(MPa)

0 0.00 0.00

2 16.33 16.82

4 19.03 20.40

7 20.69 22.73

Table 4.27: Pullout Force and Compressive Strength for 35% FA-A

Block

Age

(Days)

Pullout Force

(kN)

Estimated Strength

(MPa)

0 0.00 0.00

2 10.75 9.16

4 11.09 9.51

7 13.64 12.49

66

Table 4.28: Pullout Force and Compressive Strength for 50% FA-A

Block

Age

(Days)

Pullout Force

(kN)

Estimated Strength

(MPa)

0 0.00 0.00

2 16.60 14.85

4 17.50 15.91

7 18.21 16.75

Table 4.29: Pullout Force and Compressive Strength for 50% FA-A Slab

Age

(Days)

Pullout Force

(kN)

Estimated Strength

(Mpa)

0 0.00 0.00

2 11.79 9.50

4 14.83 12.78

7 17.10 15.44

Table 4.30: Pullout Force and Compressive Strength for 35% FA-C

Block

Age

(Days)

Pullout Force

(kN)

Estimated Strength

(Mpa)

0 0.00 0.00

2 20.35 27.48

4 22.99 35.44

7 24.28 39.82

67

4.4 Field Curing

Concrete cylinders were also field cured to compare how the field

cured values compare against the match cured values and strength

estimated from pullout forces (Upadhyaya 2008). The field cured

cylinders did not simulate the temperature profile in the concrete

blocks. Field curing represents traditional method of determining in

place compressive strength of concrete. Field cured cylinders were

tested at 2, 4 and 7 days. At each age, three cylinders were broken

and the average of the three was reported as the compressive strength.

The results are provided below.

Table 4.31: Field Cured Strength for Control Mix Block

Age

(Days)

Field Cured Strength

(MPa)

0 0.00

2 8.60

4 13.90

7 18.00

68

Table 4.32: Field Cured Strength for Control Mix Slab

Age

(Days)

Field Cured Strength

(MPa)

0 0.00

2 11.80

4 15.80

7 21.70

Table 4.33: Field Cured Strength for 35% FA-A Mix Block

Age

(Days)

Field Cured Strength

(MPa)

0 0.00

2 5.60

4 9.50

7 11.90

Table 4.34: Field Cured Strength for 50% FA-A Mix Block

Age

(Days)

Field Cured Strength

(MPa)

0 0.00

2 8.00

4 15.30

7 17.90

69

Table 4.35: Field Cured Strength for 50% FA-A Mix Slab

Age

(Days)

Field Cured Strength

(MPa)

0 0.00

2 8.70

4 14.90

7 17.10

Table 4.36: Field Cured Strength for 35% FA-C Mix Block

Age

(Days)

Field Cured Strength

(MPa)

0 0.00

2 12.00

4 20.70

7 25.50

70

Chapter 5 Exploring Alternative Approaches for Maturity Modeling

The Pull-out testing is a reliable method for estimating in place

compressive strength. The pullout tests were performed on all

concrete mixtures at the age of 2, 4 and 7 days. Using this method the

in place compressive strength was estimated for each concrete

mixture using the correlations of pullout force vs compressive

strength given in Figures 4.17 to 4.20. Testing was conducted on two

type of structural elements, concrete blocks and slabs. Slabs were

prepared for the control and the 50% FA-A mixture. To take into

account the mass concrete effects on maturity prevalent in large

(mass) concrete structures, it was important to convert the actual ages

into equivalent ages @ 23oC. The activation energy as computed in

Table 4.3 was used to perform this conversion. From the equivalent

ages, the in place compressive strength was predicted for each

mixture using the models shown in Figures 4.5 to 4.8. The pullout

force-compressive strength relationship developed in this research

was based on the 14-day strength data. In the previous study

(Upadhyaya, 2008), this relationship was based on the 28-day

strength data. Table 5.1 compares the results from both approaches

71

along with the match cured strength. Table 5.2 presents the error from

the 14-day and the 28-day pullout correlations in relation to the match

cured strength.

For the control mix, there is only a small difference between

predicted values from 14-day pullout correlations and 28-day pullout

correlations. For example, in the case of the control mix block at age

of 2 days, the 14-day pullout correlation predict a strength of 18.0

MPa while the 28-day pullout correlation predict it to be 17.90 MPa.

For the control mix block at the age of 2 days, the 14-day pullout

correlations under-predict the strength by 1.40% while the 28-day

pullout under-predict by 1.50%. In the case of the control mix slab at

the age of 4 days, the 14-day pullout based prediction under-predicts

strength by 4.60% while the 28-day based prediction under-predict by

4.80%. The 14-day pullout correlation offer small improvement over

the 28-day pullout correlation.

For the 35% FA-A mix at the ages of 2, 4 and 7 days, the strength

predicted by the 14-day pullout correlation is 9.16 MPa, 9.51 MPa

72

and 12.49 MPa while the 28-day pullout correlation predict a strength

of 9.10 MPa, 9.50 MPa and 12.40 MPa, respectively. Thus, the

difference in predicted values is very small. Overall, the 14-day

pullout under-predicts by 3.24%, 7.39% and 6.71% at the ages of 2, 4

and 7 days respectively while the 28-day pullout under-predicts the

corresponding values by 3.30%, 7.40% and 6.80%. As observed in

the case of the control mix, the 14-day pullout correlation offer small

improvement over the 28-day pullout correlation.

This is also the case with the 50% FA-A mix. The under-prediction

for the concrete block is 3.60% by both the 14-day and 28-day

pullout correlations at the age of 4 days, while for the slab the 14-day

pullout correlation under-predicts by 0.80% at the age of 2 days, and

the 28-day pullout correlation under-predicts the corresponding value

by 1.00%.

However, for the 35% FA-C mix, the 14-day pullout correlation over-

predict in-place strength at every testing age while the 28-day pullout

correlation under-predict the in-place testing age at every testing age.

73

Table 5.1: Comparison of Match Cure Strength with Predicted Strength

from the 14-day and the 28-day Pullout Correlations

Mixture Concrete

Element

Actual

Days

Equivalent

Age @ 23oC

Match

Cure

Strength

(MPa)

Strength

Predicted by 28-

Day Pullout

Correlation

(MPa)

Strength

Predicted by 14-

day Pullout

Correlation

(MPa)

Control

Mix

Block

2 3.43 19.40 17.90 18.00

4 5.99 23.70 19.60 19.73

7 8.38 26.60 23.40 23.84

Slab

2 1.72 19.50 16.80 16.82

4 2.88 25.0 20.20 20.40

7 4.66 29.60 22.30 22.73

35 % FA-A Mix

Block

2 2.40 12.40 9.10 9.16

4 4.60 16.90 9.50 9.51

7 7.19 19.20 12.40 12.49

50%

FA-A

Mix

Block

2 2.42 14.90 14.80 14.85

4 4.59 19.50 15.90 15.91

7 7.19 22.40 16.80 16.75

Slab

2 1.58 10.30 9.30 9.50

4 3.17 15.60 12.70 12.78

7 5.02 17.60 15.40 15.43

35 %

FA-C

Mix

Block

2 3.42 23.60 22.90 27.47

4 6.02 30.40 26.60 35.44

7 9.21 34.20 28.40 39.82

74

Table 5.2: Percent strength prediction differences between match cured

versus the 14-day and the 28-day Pullout correlations

Mixture Concrete

Element

Actual

Days

28-day Pullout

Correlation (%)

14-day Pullout

Correlation (%)

Control

Mix

Block

2 -1.50 -1.40

4 -4.10 -3.97

7 -3.20 -2.76

Slab

2 -2.70 -2.68

4 -4.80 -4.60

7 -7.30 -6.87

35 %

FA-A

Mix

Block

2 -0.10 -3.24

4 -3.60 -7.39

7 -6.80 -6.71

50 %

FA-A

Mix

Block

2 -0.10 -0.05

4 -3.60 -3.59

7 -5.60 -5.65

Slab

2 -1.00 -0.80

4 -2.90 -2.82

7 -2.20 -2.17

35 %

FA-C

Mix

Block

2 -0.70 3.87

4 -3.80 5.04

7 -5.80 5.62

It is also important to compare how the 14-day based maturity model

predict the in-place strengths against the in-place strength estimated

75

by the 14-day pullout correlations. Table 5.3 provides this

comparison.

For the control concrete mix block, at the age of 2 days, the strength

estimated by the pullout correlation is 18.00 MPa while the strength

predicted from the 14-day maturity model is 15.99 MPa. The strength

estimated at 4 days is practically the same by both methods. At 7 days

the pullout correlation estimates a strength of 23.84 MPa while the

14-day model estimate 21.44 MPa. Thus, as compared to the values

from the pullout correlation, the 14-day maturity models under-

predict by 11.67% at the age of 2 days, 0.68% at age of 4 days and

10.00% at age of 7 days. There is a wide variability in predicting the

in place strength of the control mix slab. At the ages of 2, 4 and 7

days, the strength predicted from 14-day maturity models are 11.20

MPa, 14.78 MPa and 18.04 MPa respectively while the

corresponding values from pullout correlations are 16.82 MPa, 20.40

MPa and 22.73 MPa. Thus, when compared with values from the

pullout correlation, the 14-day maturity models under-predict by

33.40% at age of 2 days, 27.50% at age of 4 days and 20.63% at age

of 7 days. Again, the 14-day maturity models consistently predict

lower values of in-place compressive strength for the control mix.

76

The level of under-predicting concrete strength is much higher for the

case of concrete slab. This has to do with the lack of mass effects on

maturity on slabs versus concrete blocks. .

In the case of the 35% FA-A block, the strength predicted from the

14-day maturity models are 7.98 MPa, 11.72 MPa and 14.38 MPa for

the ages of 2, 4 and 7 days respectively while the corresponding

values from the pullout correlation are 9.16 MPa, 9.51 MPa and

12.49 MPa. The values from the 14-day maturity models are more

accurate for these age groups. This is because the strength at these

ages are 12.40 MPa, 16.90 MPa and 19.20 MPa (Table 5.1) and

hence, except at the age of 2 days, the 14-day pullout correlations

have given lower estimates of in-place strength The 14-day maturity

models have estimated higher strength at age of 4 and 7 days thereby

improving the accuracy of prediction.

For the 50% FA-A slab mix results, the 14-day based maturity

method consistently predicts higher values of in-place compressive

strength at all ages than the pullout testing. This is reflected in the

case of the 50% FA-A block as well except at the age of 2 days. For

the block, the strength predicted at 4 days by the 14-day maturity

model is 18.43 MPa which is 15.82% higher than the 15.91 MPa

77

estimated strength by the pullout test. At 7 days, the strength

predicted from the 14-day maturity is 20.93 MPa which is 25% more

than the 16.75 MPa estimated by the pullout test. For the slab, at ages

of 2, 4 and 7 days, the strength estimated by the 14-day models are

10.39 MPa (9.4% more than the pullout test), 15.55 MPa (21.65%

more than the pullout test) and 19.04MPa (23.40 % more than the

pullout test) respectively. The match cured strength for 50% FA-A

block at ages 2, 4 and 7 can be found from table 5.1 and they are

14.90 MPa, 19.50 MPa and 22.40 MPa respectively. For slab, the

strength at 2, 4 and 7 days are 10.30 MPa, 15.60 MPa and 17.60

MPa. Thus, the 14-day maturity models have been successful in

reducing the degree of under-prediction in 50% FA-A mix.

In the case of the 35% FA-C block, the 14-day based maturity models

consistently under predict the in place compressive strength. The 14-

day maturity model under predict strength by 35% at age of 2 and 4

days and 33% at 7 days when compared with the results from the

pullout tests.

78

Table 5.3: Strength Prediction Comparison between Maturity Model

based on 14-Day and Pullout Test

Mixture Concrete

Element

Actual

Days

Equivalent Age @ 23oC

Strength Predicted

by 14-day Maturity

Method (MPa)

Pullout

Load

(kN)

Strength Predicted

by 14-day Pullout

Correlation (MPa)

Control

Mix

Block

2 3.43 15.99 17.20 18.00

4 5.99 19.60 18.50 19.73

7 8.38 21.44 21.50 23.84

Slab

2 1.72 11.20 16.30 16.82

4 2.88 14.78 19.0 20.40

7 4.66 18.04 20.70 22.73

35 %

FA-A

Mix

Block

2 2.40 7.98 10.80 9.16

4 4.60 11.72 11.10 9.51

7 7.19 14.38 13.6- 12.49

50 %

FA-A

Mix

Block

2 2.42 13.48 16.60 14.85

4 4.59 18.43 17.50 15.91

7 7.19 20.93 18.20 16.75

Slab

2 1.58 10.39 11.80 9.50

4 3.17 15.55 14.80 12.78

7 5.02 19.04 17.10 15.43

35 %

FA-C

Mix

Block

2 3.42 17.73 20.40 27.47

4 6.02 22.97 23.0 35.44

7 9.21 26.54 24.30 39.82

79

Table 5.4: Percent Difference between in predicted strength between 14-

day pullout and 14-day maturity method

Mixture Concrete

Element

Actual

Days

Equivalent

Age @ 23oC

14-day

Maturity

Method (%)

Control

Mix

Block

2 3.43 -11.67

4 5.99 -0.67

7 8.38 -10.07

Slab

2 1.72 -33.40

4 2.88 -27.57

7 4.66 -20.63

35 %

FA-A

Mix

Block

2 2.40 -12.90

4 4.60 23.21

7 7.19 15.16

50 %

FA-A

Mix

Block

2 2.42 -9.23

4 4.59 15.85

7 7.19 24.97

Slab

2 1.58 9.40

4 3.17 21.65

7 5.02 23.40

35 %

FA-C

Mix

Block

2 3.42 -35.45

4 6.02 -35.17

7 9.21 -33.36

80

In match curing, the temperature profile inside the structure is

recorded and replicated on testing cylinders. Thus, the cylinders

experience the same rate of strength gain as the concrete in the

structure. This makes the match cured strength predictions accurate

for the in-place compressive strength determination. Hence, it is also

important to compare the 14-day based maturity models with the

match cured strength, as well as against existing methods like the

field cured strength and the 28-day maturity models. Table 5.5

provides this comparison. Table 5.5 also provides the comparison of

setting time approach with other methods.

As can be seen from Table 5.5, the field cured strength method

predicts a strength of 8.60 MPa, 13.90 MPa and 18.0 MPa for the

control mix block at ages of 2, 4 and 7 days respectively. The match

cured strength at these ages is 19.40 MPa, 23.70 MPa and 26.60 MPa

respectively. This translates to an under prediction of 55.00%,

41.40% and 32.00% at ages of 2, 4 and 7 days respectively. Similar

effects are observed in the case of concrete slabs with the control and

the 35% FA-A mix, as well as the remaining mixtures. Thus, the field

cured method provides consistently lower strength predictions.

81

For the control mix, there is a small difference between the values

predicted from the 14-day and the 28-day based maturity models for

all the ages, as it can be seen in Table 5.5. For example, in the block

with the control mix at the age of 2 days, the 14-day maturity model

predicts a strength of 15.99 MPa while the 28-day maturity model

predicts a strength of 16.30 MPa. Even in the case of slab, at the age

of 2 days, the 14-day maturity model predicts a strength of 11.20

MPa while the 28-day maturity model predicts a strength of 11.10

MPa. Thus, these two maturity modeling methods provide similar

strength predictions at early ages. However, at the age of 28-days

for the concrete block, the 28-day based maturity model predict the

strength of 29.73 MPa while the14-day maturity model predicts a

26.40 MPa strength. The actual strength of concrete at 28 days is

29.84 MPa as determined from the match cured strength.

Differences in predictions, were examined on the 14-day and 28-day

maturity model and the results are provided in Tables 5.3 and 5.4.

This predictions may be positive or negative depending on whether

the model over-predicts or under-predicts the in-place strength at a

specific age in relation to the match cure strength. In this way, the

differences, reported here in as “errors” across all testing ages are

82

computed and an “average error” for the model is determined. This

analysis reveals that the 28-day maturity model provides closer

strength predictions to the match cured strength for the control mix. .

Both the 14-day maturity and 28-day maturity models under predict

the in place compressive strength when compared with the match

cured strength.

For the 35% FA-A block, the results are similar to the one observed

for the control mix. There is a small difference in predicted values

from the 14-day maturity and the 28-day maturity methods For

example, as it can be seen from Table 5.5, at the age of 4 days, the

14-day based maturity model predict a strength of 11.72 MPa while

the 28-day model predicts a strength of 11.20 MPa. The match cure

strength at this age was 16.90 MPa. These analyses are provided in

Tables 5.7-5.12. From the “error” analysis, it was observed that the

28-day model under-predicts the 2 day strength by 45.00% while the

14-day model under-predicts by 40.00%. The 14-day maturity model

offer some improvement over the existing 28-day maturity model at

all ages by reducing the strength under-prediction at each testing age

83

In the case of 50% FA-A mix, the 14-day model provide better

estimates of the in place strength than the 28-day based maturity

model. For example, at the age of 4 days for the concrete block, the

14-day based maturity model predicts the strength of 18.43 MPa

while for the 28-day model is 16.70 MPa. The match cure strength at

this age is 19.50 MPa. Even for the slab, the 14-day model predicts an

in-place strength of 15.54 MPa while the 28-day model predict a

strength of 14.00 MPa. The strength at this age is 15.60 MPa. The 14-

day models consistently predict higher values of in place strength at

each testing age than 28-day models thereby reducing the degree of

under-prediction. .

In the case of the 35% FA-C mix, the 28-day model consistently

predicts higher values of the in-place strength than the 14-day model.

For example at the age of 2, 4 and 7 days the 28-day model predicts

strengths of 20.70 MPa, 25.80 MPa and 29.00 MPa respectively

while the 14-day model predict strengths of 17.70 MPa, 22.97 and

26.53. The match cure strength at these ages are 23.60 MPa, 30.40

MPa and 34.20 MPa respectively. Such results perhaps imply that the

14-day model is not able to capture the mass concrete effect in large

structures as effectively as the 28-day model has. Also, from table 5.5

84

it is evident that the results of setting time approach agree with the

results from 14-day maturity approach for all the concrete mixtures.

85

Table 5.5: Strength Comparison by Different Methods

Mixture Concrete

Element

Actual

Days

Match

Cured

Strength

(MPa)

Strength

Predicted

by 14-day

Maturity

Method

(MPa)

Field

Cured

Strength

(MPa)

Strength

Predicted

by Pullout

Correlation

(MPa)

Strength

Predicted by

28-Day

Maturity

Method

(MPa)

Strength

Predicted by

14-Day

Setting Time

method

(MPa)

Control

Mix

Block

2 19.40 15.99 8.60 18.00 16.30 16.44

4 23.70 19.60 13.90 19.73 20.40 19.91

7 26.60 21.44 18.00 23.84 22.70 21.61

Slab

2 19.50 11.20 11.80 16.82 11.10 11.22

4 25.00 14.77 15.80 20.40 14.20 14.68

7 29.60 18.04 22.30 22.73 17.60 17.91

35% FA-A

Mix

Block

2 12.40 7.98 9.10 9.16 9.10 7.48

4 16.90 11.72 9.50 9.51 11.20 11.25

7 19.20 14.38 12.40 12.49 14.00 14.21

50% FA-A

Mix

Block

2 14.90 13.48 14.80 14.85 12.10 13.17

4 19.50 18.43 15.90 15.91 16.70 18.14

7 22.40 20.93 16.80 16.75 19.90 20.73

Slab

2 10.30 10.39 9.30 9.50 10.00 10.24

4 15.60 15.55 12.70 12.78 14.00 15.37

7 17.60 19.04 15.40 15.43 17.70 18.91

35% FA-C

Mix

Block

2 23.60 17.73 22.90 27.47 20.70 18.73

4 30.40 22.97 26.60 35.44 25.80 23.75

7 34.20 26.54 28.40 39.82 29.00 27.04

86

Table 5.6: Percent differences between match cured and other methods

Figure 5.1 presents strength predictions for the concrete block with

the control mix along with comparing results from field cured, match

cured, pullout strength and 28-day maturity methods. As discussed

Mixture Concrete

Element

Actual

Days

14- Day Maturity

Method (%)

Field Cure Strength

(%)

Pullout Correlation

(%)

28-Day Maturity

Method (%)

14-Day Setting Time Method (%)

Control Mix

Block

2 -17.60 -55.70 -7.20 -16.00 -15.26

4 -17.31 -41.40 -16.80 -13.90 -15.99

7 -19.40 -32.30 -10.40 -14.70 -18.76

Slab

2 -42.60 -39.50 -13.74 -43.10 -42.46

4 -40.90 -36.80 -18.40 -43.20 -41.28

7 -39.10 -24.70 -23.20 -40.50 -39.49

35% FA-A Mix

Block

2 -35.70 -26.60 -26.10 -26.60 -39.68

4 -30.70 -43.80 -43.70 -33.70 -33.42

7 -25.10 -35.40 -35.00 -27.10 -25.99

50% FA-A Mix

Block

2 -9.50 -0.67 -0.34 -18.80 -11.61

4 -5.50 -18.50 -18.40 -14.40 -6.97

7 -6.60 -25.00 -25.20 -11.16 -7.46

Slab

2 0.90 -9.70 -7.80 -2.90 -0.58

4 -0.34 -18.6 -18.10 -10.30 -1.47

7 8.20 -12.50 -12.30 0.60 7.44

35% FA-C Mix

Block

2 -24.90 -2.96 16.40 -12.30 -20.64

4 -24.40 -12.50 16.60 -15.10 -21.88

7 -22.40 -17.00 16.40 -15.20 -20.91

87

previously, the field cured method under-predicts concrete strength.

There is a small difference between the predicted values from the14-

day maturity and 28-day maturity models. However, the 28-day

maturity prediction model provides closer predictions to the 28 day

strength of concrete from the match cure approach.

Figure 5.1: Strength Comparison for Control Mix Block by different methods

Note: 28-day match cured strength predicted based on section 4.2.1 procedure

Figure 5.2 presents the strength prediction of concrete slab with the

control mix along with comparing results from the field cured, match

88

cured, pullout strength and 14 and 28-day maturity methods. Similar

effects as in the case of the control mix concrete block are observed

as well.

Figure 5.2: Strength Comparison for Control Mix Slab by different methods

Note: 28-day match cured strength predicted based on section 4.2.1 procedure

Figure 5.3 presents the strength prediction of the concrete block with

the 35% FA-A mix along with comparing results from the field cured,

89

match cured, pullout strength and the 14-day and 28-day maturity

methods. From Figure 5.3, it may appear as if the 28-day model

predict better the in-place match cure strength. However, as

explained earlier, the error analysis revealed that overall the 14-day

model provides overall a lower difference in predicted values from

match cure data than the 28-day model.

Figure 5.3: Strength Comparison for 35% FA-A Block by different methods

Note: 28-day match cured strength predicted based on section 4.2.1 procedure

90

Figure 5.4 presents the strength prediction for the concrete block of

the 50% FA-A mix along with comparing the results from the field

cured, match cured, pullout strength, and 14 and 28-day maturity

methods. At each testing age, the 14-day model predicts more

accurate the in-place strength which was also observed from the data

in Table 5.5.

Figure 5.4: Strength Comparison for 50% FA-A Block by different methods

Note: 28-day match cured strength predicted based on section 4.2.1 procedure

91

Figure 5.5 presents the strength prediction of the concrete slab with of

the 50% FA-A mix. The trend is similar to that observed in the case

of the 50% FA-A block. At each testing age, the 14-day model

predicts as well better the in-place strength, as it was observed from

Table 5.5 as well.

Figure 5.5: Strength Comparison for 50% FA-A Slab by different methods

Note: 28-day match cured strength predicted based on section 4.2.1 procedure

92

Figure 5.6 presents the strength prediction of the concrete block with

the 35% FA-C mix as discussed earlier, in this case the 28-day

maturity model provide closer in-place compressive strength than the

14-day maturity model.

Figure 5.6: Strength Comparison for 35% FA-C Block by different methods

Note: 28-day match cured strength predicted using section 4.2.1 procedure

The “error” analysis described previously for all the mixtures are

provided in Tables 5.7 to 5.12.

93

Table 5.7: Error Analysis for the Control Mix Block

Control Mix Block Variable Su Approach

Days Match Cured

Strength

% Error 14

day Model

% Error 28 day

Model

0 0.00 0.00 0.00

2 19.33 -17.26 -14.65

4 23.85 -17.84 -12.71

7 26.51 -19.13 -12.74

28 29.84 -11.52 0.33

Table 5.8: Error Analysis for the Control Mix Slab

Control Mix Slab Variable Su Approach

Days Match Cured

Strength

% Error 14

day Model

% Error 28 day

Model

0 0.00 0.00 0.00

2 19.29 -41.93 -42.33

4 25.40 -41.83 -40.58

7 29.39 -38.61 -35.61

28 34.86 -28.36 -20.31

Table 5.9: Error Analysis for the 35% FA-A Mix Block

35% FA-A Block Variable Su Approach

Days Match Cured

Strength

% Error 14

day Model

% Error 28 day

Model

0 0.00 0.00 0.00

2 12.54 -36.38 -44.10

4 16.63 -29.55 -33.38

7 19.34 -25.62 -25.60

28 23.09 -9.77 4.88

94

Table 5.10: Error Analysis for the 50% FA-A Mix Block

50% FA-A Block Variable Su Approach

Days Match Cured

Strength

% Error 14

day Model

% Error 28 day

Model

0 0.00 0.00 0.00

2 14.92 -9.64 -19.64

4 19.48 -5.39 -8.56

7 22.42 -6.64 -5.65

28 26.41 6.05 23.10

Table 5.11: Error Analysis for the 50% FA-A Mix Slab

50% FA-A Slab Variable Su Approach

Days Match Cured

Strength

% Error 14

day Model

% Error 28 day

Model

0 0.00 0.00 0.00

2 10.69 -2.80 -17.65

4 14.90 4.35 -4.00

7 17.92 6.25 3.79

28 22.48 19.83 36.04

Table 5.12: Error Analysis for the 35% FA-C Mix Block

35 % FA-C Block Variable Su Approach

Days Match Cured

Strength

% Error 14

day Model

% Error 28 day

Model

0 0.00 0.00 0.00

2 23.68 -25.12 -28.02

4 30.23 -24.00 -24.41

7 34.29 -22.62 -21.16

28 39.62 -13.18 -6.56

95

The sum of squares of prediction errors from the 14-day

setting time method the 14-day maturity method and the

28 –day maturity method in relation to the match cure

strength values are presented in table 5.13. Overall the 14-

day setting time method and 14-day maturity method

consistently provide lower sum of square errors in

strength predictions than the 28-day maturity models for

all the HVFA concrete mixtures.

96

Table 5.13: Sum of squares of prediction error by 14-day setting time

method, 14-day maturity method & 28-day maturity method vs Match

cured strength

Mix Setting Time

Method

(14 Day)

Maturity

Model

(14 Day)

Maturity

Model

(28 Day)

Control Mix

Block

13.67 14.94 11.17

Control Mix

Slab

87.15 85.66 73.09

35% FA-A

Block

21.68 18.91 42.74

50% FA-A

Block

7.83 6.81 50.17

50% FA-A

Slab

14.84 14.61 62.07

35% FA-C

Block

31.84 38.35 43.95

97

Chapter 6 Summary, Conclusions and Recommendations

In this research, a new maturity strength prediction approach was

examined based on the 14-day strength data. The ability of this

approach to predict the in-place strength of concrete mixtures was

evaluated against the traditional maturity modeling suggested by

ASTM C1074 and results from in-place strength determination. To

this effect, the following observations can be summarized for the

different mixtures:

1) The 14-day Pullout force vs compressive strength correlations were

developed in this research. The results reveal that the 14-day pullout

correlations offer small improvements in estimating the in-place

compressive strength for the conventional concrete (control) mix, the

concrete mix with 35% of type F fly ash and the concrete mix with

50% of type F fly ash. In case of the concrete mix with 35% of type C

fly ash, it was found that the 28-day pullout correlations provide

better estimates of the in-place compressive strengths.

2) The comparisons of the 14-day maturity models with the14-day

pullout correlations reveal that the 14-day maturity models

98

consistently under-predict the in-place compressive strength of the

conventional concrete (control) mix and the concrete mix with 35%

of type C fly ash. For the concrete mix with 35% of type F fly ash

and the concrete mix with 50% of type F fly ash, 14-day maturity

model consistently predict higher values of the in-place compressive

strength than the 14-day pullout correlations.

3) The strength of the concrete mixtures at each testing day is revealed

by the match curing strength and these values indicate that the 14-day

maturity models provide closer estimates to in-place strength than the

14-day pullout correlations in the case of concrete mixtures with 35%

and 50% of type F fly ash. The 14-day maturity models provide more

accurate estimates of in-place strength for these mixtures than the 14-

day pullout correlations.

4) The 14-day setting time approach for predicting in-place compressive

strength was also explored in this research. For the conventional

concrete mixtures and the mixtures with fly ash of type F, the

predictions from setting time approach are identical to the predictions

from 14-day maturity models. Thus, the 14-day setting time approach

99

has shown promising results and can be used for developing

prediction models for estimating in-place strength.

5) There is not a significant difference between the predicted values of

early age strength of the control mix concrete (at ages 2, 4 and 7

days) using both the 14-day and the 28-day based maturity models.

However, the 28-day maturity model predicts the later age strength

(at 28 days) more accurately than the 14-day model. It is

recommended to use the existing 28-day model for estimating the in-

place compressive strength of the control mix concrete.

6) In the case of the 35% FA-A mix, the 14-day maturity model

improves the estimation of the in-place compressive strength. The

28-day maturity model display greater variability in predicting the in-

place strength of 35% FA-A mix.

7) The accelerated strength development in 50% FA-A mix due to mass

effects is better captured by the 14-day maturity model than the 28-

day model. The adopted research methodology has succeeded in

reducing the variability in predicting the in-place strength.

100

8) The 28-day based maturity model predict the in-place strength of the

35% FA-C concrete more accurately than the 14-day maturity model.

On the basis of the analysis and results from various methods to

estimate in-place compressive strength of concrete, the following

conclusions were obtained from this research:

1) The 14-day pullout correlations have shown improvement over the

28-day pullout correlations in the case of conventional concrete

(control) mixtures and the concrete mixture with higher percentages

of type F fly ash. If pullout correlations are to be used for estimating

in-place strength, it is recommended to use 14-day pullout

correlations for the conventional concrete mixtures and for the

concrete mixtures with higher percentages of type F fly ash.

2) The 14-day maturity models have shown an improvement in

estimating in-place compressive strength of high volume fly ash

concrete of type F.

101

3) The mass concrete effects on concrete hydration for flat structural

elements, like concrete slabs, are minimal and do not provide

contribution to a faster rate of hydration and earlier strength gain.

Thus, the developed models were not as accurately predicting field

strength as for the case of the concrete blocks. This was also observed

from the match cure data.

4) The 14-day maturity models provide more accurate estimate of in-

place compressive strength for the concrete mixture with higher

percentages of type F fly ash than 14-day pullout correlations.

5) Field cured strength, one of the most commonly used method of in-

place strength monitoring, provides the most conservative results.

Relying solely on field cured test will delay scheduling of

construction activities thereby increasing cost of construction.

6) The 14-day setting time method and the 14-day maturity method

consistently provide lower sum of squares of errors in prediction of

in-place strength for HVFA concrete of type F and type C.

102

7) The match curing approach replicates the actual temperature profile

of the concrete structures into the testing cylinders. Thus, the testing

cylinder and the concrete structure are cured at the same temperature.

For this reason, the results from match curing are more accurate than

the other methods of in-place strength determination.

8) The process of developing maturity models by keeping the limiting

strength constant was unable to represent the hydration of concrete

mixture with different types and amounts of fly ash.

9) The 14-day setting time approach was explored and presented in this

research. The results are encouraging and this approach is suitable for

developing maturity models.

10) To develop more accurate prediction models, it is imperative to

capture the activation energy of different mixtures more accurately.

Traditional methods of estimating activation energy have yielded

variable results as seen in this research. Further investigations is

needed to explore more accurate methods for determination of

activation energy.

103

REFERENCES

Carino, Nicholas J., and H. S. Lew. "The maturity method: from

theory to application." National Institute of Standards and

Technology (2001).

Carino, N.J., The Maturity Method: Theory and Application,” ASTM

J. Cement, Concrete, and Aggregates, 6 (2), 61, 1984.

Carino, N.J and Tank, R.C., “Maturity Functions for Concrete Made

with Various Cements and Admixtures, “ACI Materials Journal, Vol.

89, No.2, March-April, PP. 188-196, 1992

Upadhyaya, Sushant. Early age strength prediction for high volume

fly ash concrete using maturity modeling. ProQuest, 2008.

Upadhyaya, Sushant, Dimitrios G. Goulias, and Karthik Obla.

"Evaluation of in place strength of high-volume fly ash

concrete." The first international conference on recent advances in

concrete technology, Arlington, VA. 2007.

Goulias, Dimitrios, and Upadhyaya Sushant. "Sustainable concrete

using high volume supplementary cementitious

materials." Proceedings of the 4th WSEAS international conference

on Energy and development-environment-biomedicine. World

Scientific and Engineering Academy and Society (WSEAS), 2011.

Upadhyaya, Sushant, Dimitrios Goulias, and Karthik Obla.

"Maturity-Based Field Strength Predictions of Sustainable Concrete

Using High-Volume Fly Ash as Supplementary Cementitious

Material." Journal of Materials in Civil Engineering 27.5 (2014):

04014165.

Li, Xue Feng, et al. "Concrete Strength Estimation Using the

Maturity Method. “Advanced Materials Research 857 (2014): 35-41

Murugan, S. Bala, G. Mohan Ganesh, and A. S. Santhi. "Regression

Models for Prediction of Compressive Strength of High Volume Fly

104

Ash (HVFA) Concrete. “Arabian Journal for Science and

Engineering (2013): 1-11.

ACI Committee 228, 2003, “In-Place Methods to Estimate Concrete

Strength (ACI 228.1R-03),” American Concrete Institute, ACI

Manual of Concrete Practice.

ASTM C29/C29M, 2003 “Standard Test Method for Bulk Density

("Unit Weight") and Voids in Aggregate,” Volume 04.02.

ASTM C33, 2003 “Standard Specification for Concrete Aggregates,”

Volume 04.02.

ASTM C39/C39M, 2005. Standard Test Method for Compressive

Strength of Cylindrical Concrete Specimens,” ASTM International,

Volume 04.02.

ASTM C109/C109M, 2007. “Standard Test Method for Compressive

Strength of Hydraulic Cement Mortars (Using 2-in. or [50-mm] Cube

Specimens),” ASTM International, Volume 04.01.

ASTM C127, 2004 Standard Test Method for Density, Relative

Density (Specific Gravity) and Absorption of Coarse Aggregate,”

Volume 04.02.

ASTM C403 / C403M-08, Standard Test Method for Time of Setting

of Concrete Mixtures by Penetration Resistance

ASTM C403, 2006, “Standard Test Method for Time of Setting of

Concrete Mixtures by Penetration Resistance,” ASTM International,

Volume 04.02

ASTM C873, 2004, “Standard Test Method for Compressive

Strength of Concrete Cylinders Cast in Place in Cylindrical Molds,”

ASTM International, Volume 04.02

105

ASTM C900, 2006, “Standard Test Method for Pullout Strength of

Hardened Concrete,” ASTM International, Volume 04.02.

ASTM C1074, 2004, “Standard Practice for Estimating Concrete

Strength by the Maturity Method,” ASTM International, Volume

04.02.